Sequencing and Characterization of the xyl Operon of a Gram-Positive Bacterium, Tetragenococcus halophila

Abstract

The xyl operon of a gram-positive bacterium, Tetragenococcus halophila (previously called Pediococcus halophilus), was cloned and sequenced. The DNA was about 7.7 kb long and contained genes for a ribose binding protein and part of a ribose transporter, xylR (a putative regulatory gene), and the xyl operon, along with its regulatory region and transcription termination signal, in this order. The DNA was AT rich, the GC content being 35.8%, consistent with the GC content of this gram-positive bacterium. The xyl operon consisted of three genes, xylA, encoding a xylose isomerase, xylB, encoding a xylulose kinase, and xylE, encoding a xylose transporter, with predicted molecular weights of 49,400, 56,400, and 51,600, respectively. The deduced amino acid sequences of the XylR, XylA, XylB, and XylE proteins were similar to those of the corresponding proteins in other gram-positive and -negative bacteria, the similarities being 37 to 64%. Each polypeptide of XylB and XylE was expressed functionally in Escherichia coli. XylE transported d-xylose in a sodium ion-dependent manner, suggesting that it is the first described xylose/Na⁺ symporter. The XylR protein contained a consensus sequence for binding catabolites of glucose, such as glucose-6-phosphate, which has been discovered in glucose and fructose kinases in bacteria. Correspondingly, the regulatory region of this operon contained a putative binding site of XylR with a palindromic structure. Furthermore, it contained a consensus sequence, CRE (catabolite-responsive element), for binding CcpA (catabolite control protein A). We speculate that the transcriptional regulation of this operon resembles the regulation of catabolite-repressible operons such as the amy, lev, xyl, and gnt operons in various gram-positive bacteria. We discuss the significance of the regulation of gene expression of this operon in T. halophila.

A gram-positive bacterium, Tetragenococcus halophila (previously called Pediococcus halophilus) (7), is used to ferment soy sauce. Soy sauce tends to form browning pigments when stored for a long time because of the amino-carbonyl reaction of amino acids with aldoses, especially pentoses (1). The elimination of pentoses from soy sauce should reduce the browning pigments (1). However, pentose utilization genes are under the regulation of catabolite repression, which represses gene expression in the presence of glucose and related catabolites. As a result, pentoses tend to remain unused after the fermentation of soy sauce (1). Therefore, elucidation of the catabolite repression regulatory system in this bacterium is the necessary step to establish a strain which can use pentoses even in the presence of glucose and thus improve the shelf life of soy sauce.

In gram-negative bacteria, catabolite repression is mediated via the crr-cya signal transduction network (25). In gram-positive bacteria, cyclic AMP is not necessarily found, and catabolite repression is thought to be mediated via the HPr kinase and CcpA pathways (12, 17, 22, 25, 34). Furthermore, many catabolite repression systems are reported to be subject to inducer exclusion or inducer expulsion (8, 19, 27, 33), which prevents the uptake of or extrudes, respectively, the inducers of the metabolic systems, making regulation more strict. The entire regulatory system of catabolite repression and inducer exclusion or inducer expulsion is called catabolite control.

The utilization of d-xylose in T. halophila is strictly regulated by the presence of d-xylose as an inducer and by the presence of glucose and related catabolites (1, 2). Neither inducer exclusion nor inducer expulsion has been detected for this system (1). Some xylose metabolic systems in other bacteria are reported to be moderately repressed by glucose and its catabolites, and others are reported to be subject to inducer exclusion or expulsion in addition to catabolite repression (8–10, 13, 14, 21, 26, 33, 34, 36–38).

In this report, we sequenced the xyl operon of T. halophila and compared it with other xyl operons to obtain insights into the regulatory mechanism of this operon.

MATERIALS AND METHODS

Abbreviations.

Abbreviations used are as follows: CRE, catabolite-responsive element; CcpA, catabolite control protein A; HPr, heat-stable histidine-containing protein; RBS, ribosome binding site.

Cloning and sequencing of the DNA containing the xyl operon.

As the first step, we used two DNA sequences as the probes for cloning: 23I [GG(T/C)TCIT(T/G)IGG(T/C)TTIGG(T/C)TC(T/G/A)AT] and 17 [TTGGGG(T/C)GG(T/C)CG(T/C)GA(A/G)GG]. We synthesized probe DNAs containing IMP and corresponding to the amino acid sequences of the XylA protein of Lactobacillus pentosus (29a); 23I and 17 correspond to the amino acid sequences from positions 233 to 240 and from positions 190 to 195 of the L. pentosus XylA protein, respectively. T. halophila was cultured in 1 liter of MRS medium (Difco) at 30°C for 2 days without shaking (1). DNA from T. halophila was prepared according to a previously described method (4). The DNA was partially cut with Sau3AI, and 5- to 20-kb fragments were fractionated by agarose gel electrophoresis and then extracted from the gel (35). The DNA was inserted into a BamHI site on λ phage vector EMBL3 (35) to construct a λ phage bank. A positive clone was selected by plaque hybridization with 23I or 17 DNA end labeled with [γ-³²P]ATP as a probe (35). A SalI fragment (4.6 kb) was subcloned into a phagemid pBluescriptII KS+ SalI site, producing pBKS+2-2. The cloned DNA lacked the upstream region of the xylA gene.

As the next step, we isolated a HindIII-SalI fragment (1 kb) corresponding to the 5′ part of the xylA gene on pBKS+2-2 as the hybridization probe. T. halophila DNA was cut with HindIII, and 3- to 8-kb fragments were fractionated by centrifugation on a sucrose density gradient (35). The DNA was used to make a λ phage bank by insertion into a HindIII site on the λ ZAP Express vector. Positive clones were used to infect Escherichia coli XL1-Blue MRF′ [Δ(mcrA)183 Δ(mcrCB-hsdSMR-mrr)173 endA1 supE44 thi-1 recA1 gyrA96 relA1 lac (F′ proAB lacI^qZ ΔM15 Tn10)] and were subjected to in vivo excision with the ExAssist helper phage, producing plasmid pBKXYL, according to the manufacturer’s protocol (see below).

We determined all the nucleotide sequences for both strands of the fragments on pBKS+2-2 and pBKXYL. The methods for sequencing and computer analysis of the genes have been described elsewhere (41). A search of the homologous proteins in the EMBL and GenBank Database Libraries was performed by use of the BLAST network service (BLASTP version 1.4).

Expression of the xylA, xylB, and xylE genes in E. coli.

We subcloned the xylA, xylB, or xylE gene in E. coli expression vector pUTE500′ (Kikkoman Co. Ltd., Tokyo, Japan) (31) by inserting a PCR-amplified fragment obtained by use of DNA primers corresponding to DNA sequences at the amino-terminal region, including the appropriate NdeI restriction site at the initiation codon of the respective gene, and at the carboxyl-terminal region, including the appropriate NdeI restriction site downstream of the termination codon (the GTG initiation codons for xylB and xylE were changed to ATG). Primer sequences for xylA were 5′-GAAAGGTGGAAATCATATGGACTATTTTGAAAACGTTCC and 5′-C ATAATCCATTTCAATCACATATGCCTTCTATACCAAATAGTCATTAA GAAGCG, those for xylB were 5′-GGTATAGAAGGTACATATGATTGAAATGGATTATGTATTAGGTC and 5′-GAGTAGTAGCTACCTCATATGTTGAATTCCTTATTTAAACCGGGTCTTCAC, and those for xylE were 5′-CAA GAGGGAGGGATAGAAGACATATGAAGTCACACCCGCTTACGCTTA CTC and 5′-TTATTTTTTTCCATATGTTCTTATAACCAAGTATTTTCTAATTGTTCAAGCG; the underlined sequences correspond to the NdeI restriction site. The subcloned expression plasmids were introduced into Escherichia coli JM109 (35). The transformants were aerobically grown in LB medium (35) with 0.2 mM isopropyl-β-d-thiogalactopyranoside to the mid-exponential phase at 37°C.

Assay of XylE activity in E. coli and T. halophila.

Cells were harvested at the exponential phase and washed with B7 minimal salts medium (42) supplemented with 20 mM NaCl. The assay procedure was essentially similar to a previously reported protocol (42). d-[¹⁴C]xylose was added to the cell suspension at a final concentration of 2 μM to start the uptake reaction at 30°C. The reaction was stopped by diluting the reaction mixture (100 μl) into 5 ml of cold B7 medium plus 20 mM NaCl, the mixture was filtered on a nitrocellulose filter (pore size, 0.45 μm; Toyo Roshi Co. Ltd., Tokyo, Japan), and the filter was washed once with cold B7 medium plus 20 mM NaCl. The radioactivity on the filter was measured with a liquid scintillation counter.

When the sodium ion dependence of the xylose uptake activity in E. coli or T. halophila was examined, cells were washed and suspended in B7 minimal salts medium (42). T. halophila was cultured in M-17 medium (Difco) supplemented with 5% NaCl and 1% xylose (instead of lactose) without aeration at 30°C for 2 days. The uptake activity was assayed as described above.

Assay of XylB activity in E. coli.

Cells were harvested at the exponential phase, washed three times with 10 mM Tris-HCl (pH 7.5)–1 mM EDTA, and suspended in the same buffer (final concentration, 10 mg of protein/ml). Cells were disrupted by sonication on ice (five times for 1 min each time; Branson Sonifier 250). After removal of nondisrupted cells by centrifugation at 30,000 × g and 4°C for 10 min, the supernatant was assayed for activity.

XylB activity was assayed according to the method of Lawlis et al. (29). The reaction mixture (1 ml) contained 50 mM Tris-HCl (pH 7.5), 5 mM MgCl₂, 1 mM EDTA, 1 mM phosphoenolpyruvate, 0.5 mM ATP, 1 mM xylulose, 0.01 mM NADH, 80 U of lactate dehydrogenase, and 80 U of pyruvate kinase. The reaction was started by the addition of sample (100 μl) at 30°C. The A₃₄₀ of the reaction mixture was monitored.

Assay of XylA activity in E. coli.

Cells were harvested at the exponential phase, washed three times with 10 mM Tris-HCl (pH 7.5)–1 mM dithiothreitol–150 mM NaCl, and suspended in the same buffer (final concentration, 10 mg of protein/ml). Cell lysates were obtained as described for the assay of XylB activity.

XylA activity was assayed according to the method of Ghangas and Wilson (16). The reaction mixture (0.95 ml) contained 50 mM Tris-maleate (pH 6.6), 1 mM MnCl₂, 5 mM d-xylose, and 100 μl of cell lysate. The reaction was started by the addition of d-xylose at 37°C and stopped by the addition of 50 μl of 50% trichloroacetic acid after incubation for 5 min. The precipitate was removed by centrifugation at 10,000 × g for 5 min, and the supernatant was analyzed for xylulose by the cysteine-carbazole method (39).

Materials.

Restriction enzymes, T4 DNA ligase, and the deletion kit for kilobase sequencing were purchased from Takara Shuzo Co. (Kyoto, Japan). GIGAPACKII Plus, BcaBEST labeling kit, λ phage vector EMBL3 (35), phagemid pBluescriptII KS+, λ ZAP Express vector, ExAssist helper phage, and E. coli XL1-Blue MRF′ were purchased from Stratagene, La Jolla, Calif. PCR was done according to the manufacturer’s recommendations with Taq DNA polymerase obtained from Takara Shuzo Co. or KOD DNA polymerase obtained from Toyobo Co. Ltd., Tokyo, Japan. [γ-³²P]ATP (110 TBq/mmol) and d-[¹⁴C]xylose (3.26 GBq/mmol) were obtained from Amersham Co. Ltd., Tokyo, Japan. Lactate dehydrogenase and pyruvate kinase were obtained from Oriental Yeast Co. Ltd., Tokyo, Japan. Other chemicals were commercial products of analytical grade.

Nucleotide sequence accession number.

The nucleotide sequence discussed in this article has been deposited in the DDBJ/EMBL/GenBank nucleotide sequence databases under accession number AB009593.

RESULTS

Sequence of the DNA.

Nearly complete portions of the inserts in pBKS+2-2 and in pBKXYL were sequenced. The G+C content for the entire 7.7-kb fragment was 35.8%, in good agreement with the value determined previously for the T. halophila genome (15). Figure 1 shows the gene organization of the xyl operon in our clone and the nucleotide sequence of its regulatory region.

FIG. 1.

Gene organization of the xyl operon of T. halophila and putative regulatory elements in its regulatory region. In the top line, the genes are indicated by arrows, which are not drawn to scale. The directions of transcription indicated by the arrows are putative. The noncoding regions following the xylE and xylR genes contained sequences with a potential structure in the mRNA that was similar to the rho-independent transcription termination signal of E. coli; we think that they are the termination signals (indicated by mushroom-like marks). Most open reading frames contained an ATG start codon, except for xylB and xylE, which started with a GTG codon and were preceded by possible RBSs at distances that fell within a range of 5 to 12 nucleotides. The assigned initiation codons are putative. The termination codons were TAG for xylA, xylR, and rbsB and TAA for xylB, xylE, and rbsC. The intergenic noncoding regions in the xyl cluster were 12 bp between xylA and xylB and 319 bp between xylB and xylE. They did not seem to form any secondary structures in the mRNA. The line drawn below the genes represents an enlarged outline of the regulatory elements in the regulatory region between xylR (left) and xylABE (right). The black boxes indicate the −35 and −10 regions for the xylR and xylABE promoters. Arrows indicate the putative operator with a palindromic sequence. In the bottom lines, the nucleotide sequence of the regulatory region for xylABE is shown. ATG in white letters indicates the location of the start codon for xylA. A cis-acting CRE is indicated by a double-headed arrow. Broken arrows indicate the putative operator with a palindrome. Putative −35 and −10 sequences are underlined. An RBS is also underlined.

We found a gene cluster consisting of three genes, designated xylA, xylB, and xylE (Fig. 1); xylA, xylB, and xylE code for a xylose isomerase, a xylulose kinase, and a xylose/Na⁺ symporter, respectively. All genes were in the same direction. There was no potential open reading frame immediately upstream of the 5′ end of xylA or downstream of the 3′ end of xylE in this direction; the minimum size for a putative open reading frame used to analyze the sequences was 40 amino acids. The xyl gene cluster seems to form an operon (Fig. 1). Upstream of xylA, there was an open reading frame in the opposite direction. We believe that it is a regulatory gene of the xyl operon, xylR (Fig. 1). Further upstream of xylA, there were two open reading frames in the same direction as the xyl operon. The deduced amino acid sequences resembled those of a ribose binding protein gene, rbsB, and a ribose transporter gene, rbsC (Fig. 1) (5, 18, 40).

Figure 1 also shows the nucleotide sequence upstream of the 5′ end of the xylA gene. A potential “core” sequence of the RBS for enterococcal genes is GGAGG (41); members of the genus Enterococcus are gram-positive bacteria having DNA and proteins showing high homology to those of T. halophila, such as the gene for HPr (3; unpublished data). A possible RBS sequence (GGTGG) preceded the ATG start codon of the xylA gene. Sequences of so-called −35 and −10 (or Pribnow) boxes (9) were found (Fig. 1). A reading frame designated xylR, which is thought to function as the repressor for the xyl operon, started at 367 bp upstream of the xylA gene in the complementary sequence, in which the potential RBS sequence and the sequences of −35 and −10 boxes were also observed (Fig. 1). There was a typical palindromic sequence in the 5′-untranslated region upstream of the xylA gene (Fig. 1). Since this xylose isomerase is an inducible enzyme whose amount is regulated by the addition of d-xylose (1), the palindromic sequence may be involved in the regulation of gene expression by d-xylose. Overlapping the −35 sequence, there was a consensus sequence for CRE (Fig. 1), which has been found for catabolite-repressible genes in other gram-positive bacteria and to which CcpA is thought to bind (17, 20, 22, 23, 34).

Primary amino acid sequences of the xyl gene products.

Searches of current protein databases with the BLAST network service revealed that several xyl gene products showed significant homologies with various xylose metabolic enzymes (Table 1).

TABLE 1.

Amino acid sequence similarity between T. halophila xyl gene products and those of other bacteria

Bacterium (DDBJ/EMBL/ GenBank accession no.)	% Identity (% similarity) to T. halophila
	XylA	XylB	XylR
L. pentosus (M57384)	48 (63)	49 (57)	33 (45)
B. subtilis (U66480)	54 (66)	NDa	39 (48)
S. xylosus (X57599)	49 (64)	48 (58)	35 (46)
E. coli (U00039)	44 (55)	37 (48)	27 (37)

^aND, Data not available. 

The xylA, xylB, xylE, and xylR gene products were predicted to be composed of 435, 502, 474, and 386 amino acids and to have molecular weights of 49,400, 56,400, 51,600, and 43,400, respectively. The sequences for XylA, XylB, and XylR were homologous to those of other xylose isomerases, xylulose kinases, and xylose repressors from various bacteria (Table 1). Furthermore, the homology search for XylR revealed sequences similar to hexose phosphate binding site sequences of hexose kinases from Streptococcus mutans and Streptomyces coelicolor (Fig. 2).

FIG. 2.

Similarity of amino acid sequence of T. halophila XylR to those of S. mutans fructokinase (DDBJ/EMBL/GenBank accession no. D13175) and S. coelicolor glucokinase (Swiss-Prot accession no. P40184). Asterisks indicate identical amino acid residues. Numbers for each line are those of amino acid residues starting from the amino terminus. Dashes indicate gaps.

Recently, a gene for a xylose transporter (xylT) was first reported for a gram-positive bacterium, Bacillus megaterium (38). The amino acid sequence alignment showed that the residues of T. halophila XylE were 51 and 64% identical to those of E. coli XylE (11) and B. megaterium XylT (38) and 60 and 81% similar, with equivalent substitutions, respectively. Hydropathy analysis by the method of Kyte and Doolittle (28) suggested that the xylE gene product encodes a hydrophobic protein with 12 membrane-spanning regions. The xylE gene is the second reported xylose transporter gene for gram-positive bacteria.

Comparison of the arrangement of the T. halophila xyl genes with those of other xyl genes.

Figure 3 shows the gene arrangements of xyl operons in bacteria. The arrangement of the genes for the T. halophila xyl operon was the same as that for the B. megaterium xyl operon (38), with the order xylR (on the reverse coding frame) and xylABE (or xylABT). Like the B. megaterium xyl operon, the T. halophila xyl operon contained the xylE (xylose transporter) gene in the operon, indicating that xylose transport activity itself is repressed in the absence of d-xylose and is subject to catabolite repression.

FIG. 3.

Organization of xyl genes of T. halophila and various bacteria. DNA sequence analysis of a 6.4-kb region revealed the presence of four open reading frames. T. halophila xylA codes for a xylose isomerase, xylB codes for a xylulose kinase, and xylE codes for a xylose transporter. xylR, transcribed in the opposite direction, codes for a xyl repressor. A, isomerase; B, kinase; R, repressor; E, transporter (for B. megaterium, T); R-T, activator; P and Q, regulatory proteins for gene expression (24). Sources for other bacteria were as follows: B. megaterium (38), L. pentosus (DDBJ/EMBL/GenBank accession no. M57384), Bacillus subtilis (DDBJ/EMBL/GenBank accession no. U66480), Streptococcus violaceoniger (29a), and E. coli (DDBJ/EMBL/GenBank accession no. U00039).

Expression of XylA, XylB, and XylE activities in E. coli.

The xylA, xylB, or xylE gene was subcloned into the cloning site, NdeI, in an E. coli expression vector (31) which has the lac promoter and operator, the lac RBS sequence, and the ATG initiation codon in the unique NdeI restriction site. The activities for XylB and XylE were successfully expressed and detected in E. coli, suggesting that the putative open reading frames and the initiation codons for these genes were probably correct.

The xylose transport activity of T. halophila was measured with B7 medium. The activity was about 0.4 nmol/min/mg of protein in the presence of 10 mM NaCl. The activity depended on the presence of NaCl (Fig. 4) but was not affected by the presence of KCl (data not shown). The apparent affinity for Na⁺ of the transport activity was estimated to be about 10 mM from Hanes-Woolf plots of the activity versus NaCl concentration. This is the first report of an Na⁺-dependent xylose transport system in bacteria. The xylose transport activity expressed in E. coli cells was 0.11 nmol/min/mg of protein after induction with 0.2 mM isopropyl-β-thiogalactoside; the background activity was 0.01 nmol/min/mg of protein. The activity expressed in E. coli showed a similar property of sodium ion dependence. Therefore, XylE of T. halophila is likely a xylose/Na⁺ symporter. It is noteworthy that the amino acid sequence of the T. halophila xylose transporter is similar to that of the E. coli xylose/H⁺ symporter (11), which is a member of the large Glut family of transporters (6, 30).

FIG. 4.

Effects of NaCl concentration on xylose uptake activity in T. halophila. The initial rate of xylose uptake in intact cells of T. halophila I-13 was measured in the presence of 0 to 70 mM NaCl and 2 μM d-[¹⁴C]xylose for 1 minute as described in Materials and Methods. A Hanes-Woolf plot of xylose uptake activity in intact cells versus outside NaCl concentration is shown. (Inset) Xylose uptake activity in intact cells versus outside NaCl concentration.

XylB activity measured with E. coli lysates was about 10.7 μmol of NADH oxidized/min/mg of protein; the background activity was negligible (below 0.01 μmol of NADH oxidized/min/mg of protein).

XylA activity measured with E. coli lysates was not different from the background activity; we concluded that XylA activity was not functionally expressed in E. coli.

DISCUSSION

In clones pBSK+2-2 and pBKXYL we found gene clusters containing a portion of the rbs operon (rbsB and a portion of rbsC), xylR, and the xyl operon, consisting of three genes (xylA, xylB, and xylE, which code for a xylose isomerase, a xylulose kinase, and a xylose transporter, respectively). It is interesting that the genes for the ribose (a pentose) metabolic pathway were found next to the xyl operon, which is susceptible to catabolite repression (1). Intergenic regions in the xyl operon did not seem to form any secondary structures in mRNA which may work as signals for transcription termination or regulation; in the B. megaterium xyl operon, the intergenic region between xylA and xylB has been shown to work as a transcription termination signal (38). We speculate that only one mRNA is transcribed for this xylA-xylB-xylE operon. Studies of the transcriptional regulation of this xyl operon are in progress.

The deduced amino acid sequences of the Xyl gene products were homologous to those of respective gene products in other bacteria (Table 1). It is noteworthy that XylE of T. halophila, which is likely to be the first described xylose/Na⁺ symporter, is homologous to the xylose/H⁺ symporter of E. coli (11). The xylose transport activity in T. halophila was not susceptible to inducer exclusion or inducer expulsion (1). Sequences relevant to inducer exclusion in transporter proteins of gram-positive bacteria need to be elucidated. It is interesting that the carboxyl-terminal hydrophilic region in XylE of T. halophila is shorter than those of the gene products of E. coli xylE and B. megaterium xylT (11, 38), since this region was suggested to be relevant to inducer exclusion in MelB of E. coli (27) and LacS of Streptococcus thermophilus (32).

The arrangement of the genes for the xyl operon was the same as that of the B. megaterium xyl operon (38). This arrangement is relevant to the observation that the xylose transport activity in T. halophila was not susceptible to inducer exclusion but that it was strictly repressed in the absence of d-xylose and was subject to catabolite repression (1, 2).

The xylB and xylE genes were successfully expressed in E. coli, suggesting that they are open reading frames. The initiation codons that we assigned are putative and await confirmation by amino acid sequencing of purified proteins. XylA activity was not functionally expressed in E. coli. We do not think that the failure of XylA expression in E. coli was due to a change in the natural 5′ sequence at the start of xylA. Instead, although we have not yet examined the production of the XylA polypeptide by using a specific antibody, we speculate that the XylA protein expressed in E. coli is unstable and becomes inactivated.

The xylose transport activity in T. halophila was Na⁺ dependent; this finding is the first report of an Na⁺-dependent xylose transporter. The property of Na⁺ dependence suggests that the xylose transporter of T. halophila is a xylose/Na⁺ symporter; in contrast, the xylose transporter in E. coli is a xylose/H⁺ symporter (11). This Na⁺ dependence is in good accord with the fact that this bacterium is halophilic, the usual salt concentration of the medium being above 15% NaCl. The possibility that this xylose transporter can work as a xylose/H⁺ symporter at a low pH (since T. halophila is a lactic acid bacterium) is not excluded.

The amino acid sequence of XylR is interesting for two reasons. (i) It has a consensus sequence for a hexose phosphate binding site found in hexose kinases (Fig. 2), suggesting that it has a binding site for such catabolites. (ii) It has a consensus sequence for a xylose binding site, since the amino acid sequence is similar to those of other XylR proteins from various bacteria (Table 1). These characteristics have been reported for the regulatory XylR proteins of the catabolite-repressible xyl operons from Bacillus subtilis and B. megaterium (9, 10, 13, 14, 21, 26, 37, 38).

From the results of this study and other studies (9, 10, 13, 14, 17, 20–23, 26, 34, 37, 38), we have devised a working model for the regulatory network of the xyl operon in T. halophila (Fig. 5). (i) Catabolites of glucose activate the HPr kinase, and then the phosphorylated form of HPr(Ser) binds to CcpA, mediating the dimerization of CcpA (12, 34, 37). CcpA with phosphorylated HPr(Ser) and a catabolite bind to the CRE sequence and repress the expression of the xyl operon (catabolite repression). (ii) A catabolite binds to XylR, activating it to bind to the xyl operator, thus repressing the xyl operon, as reported for the xyl operon in B. megaterium (antiinducer; 37). (iii) d-Xylose itself binds to XylR, inactivating it to be released from the xyl operator, thus inducing the xyl operon. This regulatory model, in which catabolites doubly repress xyl expression via CcpA (catabolite repression) and via XylR (antiinducer), is noteworthy because xylose metabolic activity in T. halophila was strictly repressed in the presence of glucose (1). In contrast, arabinose metabolic activity was repressed only 70% by the presence of glucose (catabolite control) in this bacterium (unpublished observation).

FIG. 5.

Regulatory scheme for gene expression of the xyl operon in T. halophila. The shaded arrows indicate the putative directions of transcription. The mushroom-like marks indicate the rho-independent transcription termination signals. P’s indicate promoters for the xylR and xylABE operons. T. halophila xylA codes for a xylose isomerase, xylB codes for a xylulose kinase, and xylE codes for a xylose transporter; xylR, transcribed in the opposite direction, codes for a xylose repressor. In the absence of d-xylose, XylR binds to a xyl operator (designated O) and represses the transcription of xylABE [as indicated by (−)]. In the presence of glucose, catabolites (glucose-6-phosphate [Glc-6-P], fructose-6-phosphate, or fructose-1,6-bisphosphate [F-1,6-P]) activate repressor activity [as indicated by (+)] and the CcpA-phosphorylated HPr(Ser) [CcpA-HPr(Ser)P] complex binds to CRE; both repress the transcription of xylABE [as indicated by (−)]. In the presence of xylose and only in the absence of glucose (or related catabolites), XylR is inactivated to be released from the xyl operator, inducing the transcription of xylABE. Xylulo-5-P, xylulose-5-phosphate. For details, see the Discussion.